DESCRIPTION There is increasing interest in the use of viral vectors to deliver suicide genes such as HSV-tk to CNS tumors which in combination with systemic prodrug delivery can result in drug activation in situ and tumor destruction. The applicant suggests that virus delivery to human tumors by direct inoculation will be inadequate due to the volume of the brain and the fact that glioma cells infiltrate normal brain tissue. The proposed research is intended to investigate mechanisms to increase delivery of both anti-tumor viral vectors and anti-viral agents to tumor-bearing rat brain and normal primate brain in order to evaluate vector toxicity and increase the efficacy of gene therapeutic treatments of intracerebral tumors. Experiments in AIM 1 will assess methods to deliver AV and HSV to intracerebral xenografts of LX-1 human small cell lung carcinoma in the nude rat. The applicant will compare stereotactic inoculation and convection enhanced diffusion with transvascular delivery following osmotic blood-brain barrier disruption (BBBD) or bradykinin-induced modification of the blood-tumor barrier (BTB). A monocrystalline iron oxide nanocompound (MION) will be used as a magnetic resonance imaging agent to monitor virus delivery to the CNS. In AIM 2 nude rat intracerebral tumors inoculated with viruses that express the HSV-tk gene will be treated with the anti-viral agents GCV and ACV and compared for anti-tumor activity. Again the osmotic BBBD or bradykinin-induced BTB manipulations will be tested for increased anti-tumor activity using the drugs. Aim 3 will investigate the anti-tumor efficacy of recombinant HSV and AV mutants with the best drug regimen from aim 2. Tumor killing and vector toxicity will be assessed by comparing tumor size, virus distribution, gene expression and animal survival. In aim 4, the vector system will be tested for toxicity in primates.